Podcast Transcript Summarisation
Project description
T5 Podcast Summariser
The aim of this project is to demonstrate the capabilities of Google's T5 model that has been fine-tuned on Spotify's Podcasts Dataset for automatic podcast summarisation. Simply pass in a podcast's transcript as input to the summariser, and it would output a summary containing the gist of what the podcast is about.
T5 and Spotify Podcast Dataset
- To learn more about T5, read this blog post
- To learn more about the Dataset, read its paper
Usage
This package relies on the HuggingFace NLP Library to work so you would have to install it too:
pip install transformers
pip install t5-podcast-summariser
You can now use the package as follows:
from t5_podcast_summariser import Summariser
summariser = Summariser()
transcript = """
Full Transcript of the podcast.....
"""
summary = summariser.summarise(transcript)
print(summary)
Below is a sample summary generated on a podcast transcript available here:
This week on the podcast, we talk to Ayodeji Ogunnia (@Ayodeji_Ogunnia) about his life as a bricklayer in Baltimore. We also hear from one of the most famous people in the world, and how he came to be a successful bricklayer. This is a great story for anyone who wants to learn more about what it means to be a good bricklayer. If you like what we do, please leave us a review on Apple Podcasts! Thanks for listening!
Future Work
In the next update, I will make it possible for the top-n sentences to be extracted from the transcript before a summary is generated.
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Hashes for t5_podcast_summariser-0.0.5.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | fee1496dd23ea736dbe7fde7f6f297e21c1111bc59e80bb158698862d1028cef |
|
MD5 | 0a38f59228dcd50eb1970a8546b0c4ec |
|
BLAKE2b-256 | 92efa06501f94c8c5bfa6a7fe547b837fc62bf2fe50ca12c92af2c66f3b0b12f |